BACKGROUND: Infants born after IVF are often twins, and singleton IVF babies have an increased risk for preterm birth. Both conditions are likely to increase morbidity. We examined the frequency and duration of hospitalization required by babies born after IVF, and compared this information with all infants born in Sweden during the same time period. METHODS: We used a nationwide registration of IVF pregnancies from 1984 to 1997 and a nationwide register of all in-patient care up to the end of 1998. We identified 9056 live born infants after IVF treatment and compared them with 1 417 166 non-IVF live born infants. RESULTS: The highest odds ratio (OR ~3) was seen for neonatal hospitalization, but an increased OR (1.2–1.3) was noted for children up to 6 years of age. The OR for being hospitalized after IVF was 1.8, but when the analysis was restricted to term infants it was 1.3 and this excess was then explainable by maternal subfertility. Statistically significant increased ORs were seen for hospitalization for cerebral palsy (1.7), epilepsy (1.5), congenital malformation (1.8) or tumour (1.6), but also for asthma (1.4) or any infection (1.4). When information from the Swedish Cancer Registry was used, no excess risk for childhood cancer was found. The average number of days spent in hospital by IVF and non-IVF children was 9.5 and 3.6 respectively. CONCLUSIONS: The increased hospitalization of IVF children is, to a large extent, due to the increased incidence of multiple births. Therefore, the increased costs associated with this may be reduced by the use of single embryo transfers, with the savings in health care costs being offset against the increased number of embryo transfer cycles required to maintain the pregnancy rate.